I need to develop a 'Fault' process to handle the below.  Does anyone have any experience in setting and examining the custom JMS property?

Basically, there is a Fault Handler process. The main process will have both the Fault Endpoint and Rejected Message Endpoint pointing to the ***.Fault.Entry.
So any error will go the Fault Handler that is listening to this endpoint.
See sample Fault Handler attached.
We can do the CBR condition different ways but maybe creating a custom JMS property like ‘Retry count’ is straightforward.
CBR checks retry count on the message which will initially be set to zero on the main process.
If less than max, proceed to first step on right. This increments retry property using enrichment service and then pass to next step which calls the web service again. Web service can be successful at which point response is returned to REPLY_TO. If it fails, message goes to ***.Fault.Entry and back to this Fault process. At that point, retry is at 1 and so on…
If max retry is reached, proceed to first step on left to build an error message and pass to REPLY_TO (unless we decide to use an RME instead for a delayed and manual retry).
The main process will have
Entry Endpoint=***.Main.Entry
Exit Endpoint = REPLY_TO
Fault Endpoint= ***.Fault.Entry 
Rejected Message Endpoint=***.Fault.Entry
The fault process will have
Entry Endpoint=***.Fault.Entry 
Exit Endpoint = REPLY_TO
Fault Endpoint= ***.Fault.Entry 
Rejected Message Endpoint=***.Fault.Entry 
In addition to retry count, CBR can check for specific types of error messages that may require the response to be sent immediately rather than forcing it to loop n times.